CN106599215A - Question generation method and question generation system based on deep learning - Google Patents
Question generation method and question generation system based on deep learning Download PDFInfo
- Publication number
- CN106599215A CN106599215A CN201611168600.4A CN201611168600A CN106599215A CN 106599215 A CN106599215 A CN 106599215A CN 201611168600 A CN201611168600 A CN 201611168600A CN 106599215 A CN106599215 A CN 106599215A
- Authority
- CN
- China
- Prior art keywords
- question
- semantic
- seed
- words
- candidate
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000013135 deep learning Methods 0.000 title claims abstract description 14
- 230000011218 segmentation Effects 0.000 claims abstract description 30
- 238000001514 detection method Methods 0.000 claims abstract description 13
- 238000013145 classification model Methods 0.000 claims abstract description 12
- 241000238017 Astacoidea Species 0.000 description 10
- 241000238557 Decapoda Species 0.000 description 6
- 230000008569 process Effects 0.000 description 4
- 238000012216 screening Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 3
- 241000251468 Actinopterygii Species 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 2
- 230000009193 crawling Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 235000011194 food seasoning agent Nutrition 0.000 description 2
- 239000004615 ingredient Substances 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 235000013372 meat Nutrition 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000003058 natural language processing Methods 0.000 description 2
- 238000005215 recombination Methods 0.000 description 2
- 230000006798 recombination Effects 0.000 description 2
- 238000011895 specific detection Methods 0.000 description 2
- 230000007480 spreading Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/3331—Query processing
- G06F16/3332—Query translation
- G06F16/3335—Syntactic pre-processing, e.g. stopword elimination, stemming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/35—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Human Computer Interaction (AREA)
- Machine Translation (AREA)
Abstract
The invention provides a question generation method and question generation system based on deep learning. The question generation method comprises the following steps: acquiring a seed question; performing word segmentation on the seed question; performing sememic word extension on the seed question after the word segmentation; recombining the extended sememic words to generate candidate questions; performing semantic detection on the candidate questions through a preset semantic classification model so as to acquire the candidate question with the correct semantic. By implementation of the method disclosed by the embodiment of the invention, not only is the labor saved, but also is the accuracy of the generated question improved.
Description
Technical Field
The invention relates to the field of electric data processing, in particular to a question generation method and a question generation system based on deep learning.
Background
With the introduction of the Web2.0 era, tens of thousands of web pages are newly added on the Internet every day. In the face of massive web pages, a user can hardly find the information desired by the user quickly. Search engines represented by Google, hundredths help users to find useful information quickly and accurately by calculating the relevance of web page contents for key words input by users and then returning several web pages most relevant to user queries.
Although web search engines have achieved great success in today's numerous internet applications, they have more or less some disadvantages as the main means for people to obtain information, including the following: (1) usually, the search engine returns several web pages most relevant to the user query, and the user needs to browse the result list returned by the search engine from top to bottom one by one and see most contents of the web pages to summarize and summarize the really needed contents and information from the results list. This often takes a lot of time for the user and is prone to annoy the user because it is too cumbersome; (2) the processing of keyword queries by search engines can cause users who are not familiar with using search engines to submit different queries to the search engines many times to expect desired information, but can also obtain a lot of unnecessary information, which brings much inconvenience to the users when the users effectively utilize the search engines to help the users to retrieve the information; (3) most of the existing search engines are based on keyword query, and the input of a user needs to be subjected to word segmentation processing, so that input semantic information is usually lost, and the result returned by the search engine is not accurate enough.
To solve the above series of problems, question-answering systems have come into play. Firstly, the query mode of question answering is a complete spoken question, so that the time for elaborately constructing query conditions is saved for a user, and the semantic information of the question is fully utilized; and secondly, the return of the question-answering system is a high-precision webpage result or a definite answer string, so that the situation that the user spends more time to summarize and summarize the required content information from the whole webpage is avoided. Artificial intelligence has developed rapidly in recent years due to the successful application of machine learning in the fields of computer vision, natural language processing. The question-answering system has attracted more and more attention as an important task in the field of natural language processing. At present, a plurality of domain intelligent question-answering systems, such as intelligent customer service, are also emerged on the market. By answering the customer's questions, the intelligent customer service can greatly reduce the labor cost of the enterprise.
At the present stage, a large number of question sentences are needed to be used as training corpora for constructing the intelligent question-answering system by using a statistical machine learning method. The current commonly used question collection methods mainly comprise: manual collection and web crawlers. The number of manually collected question sentences is usually limited, and the cost is high, so that the method is not feasible for intelligent system training with large data demand. The question data crawled by the web crawler may contain a large amount of noise, and if the data containing a large amount of noise is directly taken as training data, the model obtained by training has a large problem. According to the invention, a large amount of question data can be automatically generated by manually collecting a small amount of seed data and using a deep learning method. Experiments show that the question sentence generated by the method is high in quality, and a good intelligent question-answering system can be constructed.
Disclosure of Invention
In view of this, the present invention provides a question generation method and a question generation system based on deep learning, so as to solve the problem of low accuracy of the question generated in the prior art.
Specifically, the invention is realized by the following technical scheme:
the invention provides a question generating method based on deep learning, which comprises the following steps:
acquiring a seed question;
segmenting the seed question sentence;
performing semantic word expansion on the participle question sentence;
recombining the expanded semantic words to generate candidate question sentences;
and performing semantic detection on the candidate question through a preset semantic classification model to obtain the candidate question with correct semantics.
The invention also provides a question generating system based on deep learning, which comprises the following components:
a seed question acquiring unit for acquiring a seed question;
the word segmentation unit is used for segmenting the seed question sentence;
the semantic word expansion unit is used for performing semantic word expansion on the participle question sentences;
a candidate question generating unit, configured to recombine the expanded semantic words to generate candidate question;
and the semantic detection unit is used for performing semantic detection on the candidate question through a preset semantic classification model to obtain a candidate question with correct semantics.
According to the embodiment of the invention, the seed question is obtained, the seed question is segmented, the semantic word expansion is carried out on the segmented seed question, the expanded semantic words are recombined to generate the candidate question, and the semantic detection is carried out on the candidate question through the preset semantic classification model to obtain the candidate question with correct semantics, so that not only is the labor saved, but also the accuracy of the generated question is improved.
Drawings
Fig. 1 is a flowchart of a question generation method based on deep learning according to an exemplary embodiment of the present invention;
fig. 2 is a block diagram of a question generation system based on deep learning according to an exemplary embodiment of the present invention.
Detailed Description
Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the invention, as detailed in the appended claims.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
It is to be understood that although the terms first, second, third, etc. may be used herein to describe various information, these information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present invention. The word "if" as used herein may be interpreted as "at … …" or "when … …" or "in response to a determination", depending on the context.
Fig. 1 is a flowchart of a question generation method based on deep learning according to an exemplary embodiment of the present invention, where the method includes the following steps:
step S101, a seed question is obtained.
In the embodiment of the present invention, a seed question is a question with definite semantics and correct syntax, such as: "how much weather today is Beijing? "the question contains explicit semantic information (asking for the weather condition of beijing today) and accords with normal grammatical expression, while" how much beijing weather is today ", although there is some semantic information indicating that weather is being asked, it does not accord with normal grammatical expression, and therefore, it cannot be used as a seed question. However, "what weather Beijing today" has not only semantic information missing, but also syntax expression errors, and cannot be used as a seed question.
The acquiring of the seed question sentence comprises the following steps:
1. directionally acquiring a seed question sentence according to preset keywords and question words; or,
2. the seed question was obtained by manual collection.
In the embodiment of the invention, the seed question is obtained mainly by two methods: directional grabbing and manual collection. And directionally grabbing through predefined keywords and query words, selecting a database with strong field, crawling the database, and directionally grabbing according to whether the crawled content contains the keywords and the query words. And the manual collection is to use the scene description by publishing the keywords and the question, and to acquire the seed question by a crowdsourcing mechanism.
And step S102, performing word segmentation on the seed question sentence.
In the embodiment of the invention, different from the common word segmentation only by using a general word segmentation device, the word segmentation is performed on the obtained seed question sentence in a cutting-segmenting-cutting mode, so that the requirement of general word segmentation can be met, and the effective word segmentation can be performed according to the application scene of the seed question sentence. The word segmentation method can not only ensure the integrity of the segmentation of the question sentences of the seeds, but also improve the segmentation precision of field problems and reduce the loss of semantic information.
And step S103, performing semantic word expansion on the participle question sentence.
In the embodiment of the invention, the stop words and repeated words generated in the word segmentation process are removed by the seed problem after word segmentation, and the words in the question sentence are subjected to semantic expansion through word2 vec.
For example: "braised prawn/n, has/v, those/r, do/n,? W ", the results after semantic expansion of a word by word2vec are shown in the following table:
and step S104, recombining the expanded semantic words to generate candidate question sentences.
In the embodiment of the present invention, after the semantic word is subjected to the above word segmentation expansion, the candidate sentence needs to be generated by recombination, and specifically, the candidate sentence is generated by recombining the semantic word by cartesian convolution.
The step of recombining the expanded semantic words to generate candidate question sentences comprises the following steps:
and carrying out full arrangement on the semantic words, and carrying out Cartesian convolution according to the expansion words to generate candidate question sentences.
Such as: the braised prawn (1) has (2) and (3) practices (4). And performing Cartesian convolution on the expansion words according to the positions of the full arrangement of each semantic word in 24 groups. If one of the arrangements is (1), (2), (3) and (4), the cartesian convolution generated by the spreading word is:
{ braised crayfish, braised shrimp, braised fish, cooked-back meat } x { yes, none, nothing, know what, which, what, about, ask, where } x { practices, ingredients, eating, seasoning, method }.
A candidate question generated from the cartesian convolution is: how to braised crayfish.
And step S105, performing semantic detection on the candidate question through a preset semantic classification model to obtain a candidate question with correct semantics.
In the embodiment of the present invention, in all generated question candidates, not all question candidates have semantics, such as: "what the braised crayfish does", there are also candidate sentences which contain semantics but have abnormal grammars, such as: the question of "what the braised crayfish did, and" what the braised crayfish did "is considered to be normal. Therefore, a preset semantic classification model is adopted to screen the candidate question sentences: and screening out question sentences containing semantics and screening out question sentences with correct semantics.
The specific detection process is as follows:
firstly, searching candidate semantic question in a conventional question index base to search for sentences with normal semantics related to the candidate semantic question, calculating the probability value of the normal sentences of the searched sentences with normal semantics by using a language model, and simultaneously selecting the smallest probability sentence value, meanwhile, calculating the sentence probability value of the candidate semantic question by using the language model, judging whether the sentence probability value of the candidate semantic question is greater than the minimum value of the normal sentence probability, if so, determining that the semantics of the candidate semantic question is correct, otherwise, determining that the semantics is wrong. And storing the question with correct semantics in a question bank.
According to the embodiment of the invention, the seed question is obtained, the seed question is segmented, the semantic word expansion is carried out on the segmented seed question, the expanded semantic words are recombined to generate the candidate question, and the semantic detection is carried out on the candidate question through the preset semantic classification model to obtain the candidate question with correct semantics, so that not only is the labor saved, but also the accuracy of the generated question is improved.
Fig. 2 is a structural diagram of a question generation system based on deep learning according to an exemplary embodiment of the present invention, where the question generation system includes:
a question-seed acquiring unit 201, configured to acquire a question-seed.
In the embodiment of the present invention, a seed question is a question with definite semantics and correct syntax, such as: "how much weather today is Beijing? "the question contains explicit semantic information (asking for the weather condition of beijing today) and accords with normal grammatical expression, while" how much beijing weather is today ", although there is some semantic information indicating that weather is being asked, it does not accord with normal grammatical expression, and therefore, it cannot be used as a seed question. However, "what weather Beijing today" has not only semantic information missing, but also syntax expression errors, and cannot be used as a seed question.
The question-of-seed obtaining unit 201 includes:
an orientation acquiring subunit 2011, configured to directionally acquire a seed question according to a preset keyword and a question word; or,
a manual collection subunit 2012, configured to obtain the seed question sentence through manual collection.
In the embodiment of the invention, the seed question is obtained mainly by two methods: directional grabbing and manual collection. And directionally grabbing through predefined keywords and query words, selecting a database with strong field, crawling the database, and directionally grabbing according to whether the crawled content contains the keywords and the query words. And the manual collection is to use the scene description by publishing the keywords and the question, and to acquire the seed question by a crowdsourcing mechanism.
And a word segmentation unit 202, configured to perform word segmentation on the seed question sentence.
In the embodiment of the invention, different from the common word segmentation only by using a general word segmentation device, the word segmentation is performed on the obtained seed question sentence in a cutting-segmenting-cutting mode, so that the requirement of general word segmentation can be met, and the effective word segmentation can be performed according to the application scene of the seed question sentence. The word segmentation method can not only ensure the integrity of the segmentation of the question sentences of the seeds, but also improve the segmentation precision of field problems and reduce the loss of semantic information.
And the semantic word expansion unit 203 is used for performing semantic word expansion on the participle question sentence.
In the embodiment of the invention, the stop words and repeated words generated in the word segmentation process are removed by the seed problem after word segmentation, and the words in the question sentence are subjected to semantic expansion through word2 vec.
For example: "braised prawn/n, has/v, those/r, do/n,? W ", the results after semantic expansion of a word by word2vec are shown in the following table:
and a question candidate generating unit 204, configured to recombine the expanded semantic words to generate question candidates.
In the embodiment of the present invention, after the semantic word is subjected to the above word segmentation expansion, the candidate sentence needs to be generated by recombination, and specifically, the candidate sentence is generated by recombining the semantic word by cartesian convolution.
The step of recombining the expanded semantic words to generate candidate question sentences comprises the following steps:
and carrying out full arrangement on the semantic words, and carrying out Cartesian convolution according to the expansion words to generate candidate question sentences.
Such as: the braised prawn (1) has (2) and (3) practices (4). And performing Cartesian convolution on the expansion words according to the positions of the full arrangement of each semantic word in 24 groups. If one of the arrangements is (1), (2), (3) and (4), the cartesian convolution generated by the spreading word is:
{ braised crayfish, braised shrimp, braised fish, cooked-back meat } x { yes, none, nothing, know what, which, what, about, ask, where } x { practices, ingredients, eating, seasoning, method }.
A candidate question generated from the cartesian convolution is: how to braised crayfish.
And the semantic detection unit 205 is configured to perform semantic detection on the candidate question through a preset semantic classification model, and obtain a candidate question with correct semantics.
In the embodiment of the present invention, in all generated question candidates, not all question candidates have semantics, such as: "what the braised crayfish does", there are also candidate sentences which contain semantics but have abnormal grammars, such as: the question of "what the braised crayfish did, and" what the braised crayfish did "is considered to be normal. Therefore, a preset semantic classification model is adopted to screen the candidate question sentences: and screening out question sentences containing semantics and screening out question sentences with correct semantics.
The specific detection process is as follows:
firstly, searching candidate semantic question in a conventional question index base to search for sentences with normal semantics related to the candidate semantic question, calculating the probability value of the normal sentences of the searched sentences with normal semantics by using a language model, and simultaneously selecting the smallest probability sentence value, meanwhile, calculating the sentence probability value of the candidate semantic question by using the language model, judging whether the sentence probability value of the candidate semantic question is greater than the minimum value of the normal sentence probability, if so, determining that the semantics of the candidate semantic question is correct, otherwise, determining that the semantics is wrong. And storing the question with correct semantics in a question bank.
According to the embodiment of the invention, the seed question is obtained, the seed question is segmented, the semantic word expansion is carried out on the segmented seed question, the expanded semantic words are recombined to generate the candidate question, and the semantic detection is carried out on the candidate question through the preset semantic classification model to obtain the candidate question with correct semantics, so that not only is the labor saved, but also the accuracy of the generated question is improved.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the scheme of the invention. One of ordinary skill in the art can understand and implement it without inventive effort.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.
Claims (8)
1. A question generation method based on deep learning is characterized by comprising the following steps:
acquiring a seed question;
segmenting the seed question sentence;
performing semantic word expansion on the participle question sentence;
recombining the expanded semantic words to generate candidate question sentences;
and performing semantic detection on the candidate question through a preset semantic classification model to obtain the candidate question with correct semantics.
2. The question generation method of claim 1, wherein said obtaining a seed question comprises:
directionally acquiring a seed question sentence according to preset keywords and question words; or,
the seed question was obtained by manual collection.
3. The question generation method according to claim 1, characterized in that the seed question is semantically word expanded by word2 vec.
4. The question generation method according to claim 1, wherein said recombining the expanded semantic words to generate candidate question includes:
and carrying out full arrangement on the semantic words, and carrying out Cartesian convolution according to the expansion words to generate candidate question sentences.
5. A question generation system based on deep learning, characterized by comprising:
a seed question acquiring unit for acquiring a seed question;
the word segmentation unit is used for segmenting the seed question sentence;
the semantic word expansion unit is used for performing semantic word expansion on the participle question sentences;
a candidate question generating unit, configured to recombine the expanded semantic words to generate candidate question;
and the semantic detection unit is used for performing semantic detection on the candidate question through a preset semantic classification model to obtain a candidate question with correct semantics.
6. The question generation system according to claim 5, wherein the seed question acquisition unit includes:
the directional acquisition subunit is used for directionally acquiring the seed question sentences according to preset keywords and the question words; or,
and the manual collection subunit is used for acquiring the seed question sentence through manual collection.
7. The question generation system according to claim 5, wherein the seed question is semantically word expanded by word2 vec.
8. The question generation system according to claim 5, wherein said recombining said expanded semantic words to generate candidate question includes:
and carrying out full arrangement on the semantic words, and carrying out Cartesian convolution according to the expansion words to generate candidate question sentences.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611168600.4A CN106599215A (en) | 2016-12-16 | 2016-12-16 | Question generation method and question generation system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611168600.4A CN106599215A (en) | 2016-12-16 | 2016-12-16 | Question generation method and question generation system based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106599215A true CN106599215A (en) | 2017-04-26 |
Family
ID=58599678
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611168600.4A Pending CN106599215A (en) | 2016-12-16 | 2016-12-16 | Question generation method and question generation system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106599215A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107368547A (en) * | 2017-06-28 | 2017-11-21 | 西安交通大学 | A kind of intelligent medical automatic question-answering method based on deep learning |
| CN107423363A (en) * | 2017-06-22 | 2017-12-01 | 百度在线网络技术（北京）有限公司 | Art generation method, device, equipment and storage medium based on artificial intelligence |
| CN109033390A (en) * | 2018-07-27 | 2018-12-18 | 深圳追科技有限公司 | The method and apparatus for automatically generating similar question sentence |
| WO2020007027A1 (en) * | 2018-07-04 | 2020-01-09 | 平安科技（深圳）有限公司 | Online question-answer method, apparatus, computer equipment and storage medium |
| CN111061851A (en) * | 2019-12-12 | 2020-04-24 | 中国科学院自动化研究所 | Given fact-based question generation method and system |
| CN111222309A (en) * | 2020-01-15 | 2020-06-02 | 深圳前海微众银行股份有限公司 | Question generation method and device |
| CN111444316A (en) * | 2020-03-11 | 2020-07-24 | 浙江大学 | Knowledge graph question-answer oriented composite question analysis method |
| CN113449117A (en) * | 2021-06-24 | 2021-09-28 | 武汉工程大学 | Bi-LSTM and Chinese knowledge graph-based composite question-answering method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103593335A (en) * | 2013-09-05 | 2014-02-19 | 姜赢 | Chinese semantic proofreading method based on ontology consistency verification and reasoning |
| CN103678576A (en) * | 2013-12-11 | 2014-03-26 | 华中师范大学 | Full-text retrieval system based on dynamic semantic analysis |
| CN104050256A (en) * | 2014-06-13 | 2014-09-17 | 西安蒜泥电子科技有限责任公司 | Initiative study-based questioning and answering method and questioning and answering system adopting initiative study-based questioning and answering method |
| CN105589844A (en) * | 2015-12-18 | 2016-05-18 | 北京中科汇联科技股份有限公司 | Missing semantic supplementing method for multi-round question-answering system |
-
2016
- 2016-12-16 CN CN201611168600.4A patent/CN106599215A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103593335A (en) * | 2013-09-05 | 2014-02-19 | 姜赢 | Chinese semantic proofreading method based on ontology consistency verification and reasoning |
| CN103678576A (en) * | 2013-12-11 | 2014-03-26 | 华中师范大学 | Full-text retrieval system based on dynamic semantic analysis |
| CN104050256A (en) * | 2014-06-13 | 2014-09-17 | 西安蒜泥电子科技有限责任公司 | Initiative study-based questioning and answering method and questioning and answering system adopting initiative study-based questioning and answering method |
| CN105589844A (en) * | 2015-12-18 | 2016-05-18 | 北京中科汇联科技股份有限公司 | Missing semantic supplementing method for multi-round question-answering system |
Non-Patent Citations (1)
| Title |
|---|
| 董日壮: "社区问答检索系统的设计与实现", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107423363A (en) * | 2017-06-22 | 2017-12-01 | 百度在线网络技术（北京）有限公司 | Art generation method, device, equipment and storage medium based on artificial intelligence |
| CN107423363B (en) * | 2017-06-22 | 2021-02-19 | 百度在线网络技术（北京）有限公司 | Artificial intelligence based word generation method, device, equipment and storage medium |
| CN107368547A (en) * | 2017-06-28 | 2017-11-21 | 西安交通大学 | A kind of intelligent medical automatic question-answering method based on deep learning |
| WO2020007027A1 (en) * | 2018-07-04 | 2020-01-09 | 平安科技（深圳）有限公司 | Online question-answer method, apparatus, computer equipment and storage medium |
| CN109033390A (en) * | 2018-07-27 | 2018-12-18 | 深圳追科技有限公司 | The method and apparatus for automatically generating similar question sentence |
| CN111061851A (en) * | 2019-12-12 | 2020-04-24 | 中国科学院自动化研究所 | Given fact-based question generation method and system |
| CN111061851B (en) * | 2019-12-12 | 2023-08-08 | 中国科学院自动化研究所 | Question generation method and system based on given facts |
| CN111222309A (en) * | 2020-01-15 | 2020-06-02 | 深圳前海微众银行股份有限公司 | Question generation method and device |
| CN111444316A (en) * | 2020-03-11 | 2020-07-24 | 浙江大学 | Knowledge graph question-answer oriented composite question analysis method |
| CN111444316B (en) * | 2020-03-11 | 2023-08-29 | 浙江大学 | Knowledge graph question-answering-oriented compound question analysis method |
| CN113449117A (en) * | 2021-06-24 | 2021-09-28 | 武汉工程大学 | Bi-LSTM and Chinese knowledge graph-based composite question-answering method |
| CN113449117B (en) * | 2021-06-24 | 2023-09-26 | 武汉工程大学 | Bi-LSTM and Chinese knowledge graph based compound question-answering method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106599215A (en) | Question generation method and question generation system based on deep learning | |
| JP6309644B2 (en) | Method, system, and storage medium for realizing smart question answer | |
| CN108376160B (en) | Chinese knowledge graph construction method and system | |
| CN108280114B (en) | Deep learning-based user literature reading interest analysis method | |
| Kumar et al. | Keyword query based focused Web crawler | |
| CN103678576B (en) | The text retrieval system analyzed based on dynamic semantics | |
| CN107180045B (en) | Method for extracting geographic entity relation contained in internet text | |
| US10366093B2 (en) | Query result bottom retrieval method and apparatus | |
| CN102262634B (en) | Automatic questioning and answering method and system | |
| CN109947952B (en) | Retrieval method, device, equipment and storage medium based on English knowledge graph | |
| CN111488467B (en) | Construction method and device of geographical knowledge graph, storage medium and computer equipment | |
| CN106202294B (en) | Related news computing method and device based on keyword and topic model fusion | |
| CN104199833B (en) | The clustering method and clustering apparatus of a kind of network search words | |
| CN105045875B (en) | Personalized search and device | |
| CN102955848B (en) | A kind of three-dimensional model searching system based on semanteme and method | |
| CN112667794A (en) | Intelligent question-answer matching method and system based on twin network BERT model | |
| CN105975558A (en) | Method and device for establishing statement editing model as well as method and device for automatically editing statement | |
| WO2008106667A1 (en) | Searching heterogeneous interrelated entities | |
| CN107168991A (en) | A kind of search result methods of exhibiting and device | |
| CN109948154B (en) | Character acquisition and relationship recommendation system and method based on mailbox names | |
| CN113254671B (en) | Atlas optimization method, device, equipment and medium based on query analysis | |
| US20150206101A1 (en) | System for determining infringement of copyright based on the text reference point and method thereof | |
| CN110275949A (en) | Automatic response method and system for loan application | |
| CN110413882B (en) | Information pushing method, device and equipment | |
| CN118296120A (en) | Large-scale language model retrieval enhancement generation method for multi-mode multi-scale multi-channel recall |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170426 |